Introduction

In the past decade or so, radical-driven peptide fragmentation mass spectrometry (MS) has been a subject extensively investigated by many groups worldwide because of its promising potential as another powerful tool for peptide sequencing tandem mass spectrometry [1–16]. The radical-driven fragmentation MS methods have been studied in a variety of aspects, such as the generation and migration of a radical site [3–5, 8, 11, 17], the stability of radical ions [18, 19], peptide fragmentation pathways [20–23], thermodynamics [6, 24], applications [25–27], and gas-phase structure elucidations [14, 28, 29]. Several interesting methodologies have been introduced for generating a radical site on the peptide manifold. Important examples are the collision induced (activation) dissociation (CID/CAD) of ternary metal complexes with peptide and auxiliary ligands [1, 4], the photodissociation of photolabile radical precursors [22], the photodetachment of peptide anions [16], ion-molecule reactions [10, 15], and CID of radical precursor-incorporated peptides [2, 3, 8].

Among the many methods mentioned above, our group has focused on the method based on the introduction of a radical precursor, which was originally independently developed by the Beauchamp and the Porter groups [2, 3, 8]. In this method, the radical precursor moiety [in our case, o-TEMPO-Bz-C(O)-NHS] is conjugated into the peptides of interest, particularly onto the N-terminus of peptides; TEMPO is the acronym of 2-(2,2,6,6-tetramethyl piperidine-1-oxyl) [8, 13, 14, 26, 30–36]. The radical site is readily generated through initial collisional activation (CA). Additional CA applied to the peptides with a radical site induces peptide backbone dissociations. These CAs form the basis of the free radical-initiated peptide sequencing (FRIPS) approach of the Beauchamp group [3]. It should be noted that these FRIPS experiments were mostly conducted using ion-trap mass spectrometers. Most fragment ions observed in these types of experiments were identified to occur via radical-driven peptide fragmentation pathways: dominant a/x- and c/z-type fragments.

The TEMPO-assisted FRIPS approach has a potential to be used as an alternative to electron capture/transfer dissociation for peptide identification and characterization [8, 13, 14, 26, 30–36]. However, a few limitations of the FRIPS approach have been addressed. First, the peptides of interest can be over-conjugated with the radical precursors. The o-TEMPO-Bz-C(O)-NHS reagent can undergo double conjugations with lysine-ending tryptic peptides (i.e., at the N-terminal amine and the lysine side chain [36]). It was observed that the doubly conjugated peptides would not readily lead to radical-driven peptide backbone fragmentations upon successive applications of CA. Therefore, to circumvent the double-conjugation issue, a lysine side chain blocking strategy was successfully utilized by guanidination of the lysine side chain [36].

Second, two-step CA processes are needed to achieve substantial radical-driven peptide backbone fragmentations in positive ion mode (Scheme 1) [30]. Two-step applications of CA make our TEMPO-assisted FRIPS approach less attractive as a practical tool for peptide analyses. Although a single-step CA was shown to be sufficient to result in extensive radical-driven peptide fragmentations in negative ion mode, this two-step CA issue still needs to be resolved in positive ion mode because most peptide tandem mass spectrometry experiments are conducted in positive ion mode.

Scheme 1
scheme 1

An overall experimental flow of the TEMPO-assisted FRIPS mass spectrometry. [Scheme reproduced from International Journal of Mass Spectrometry. Guanidination of Lysine Residue Improves the Sensitivity and Interpretation of Free Radical Initiated Peptide Sequencing (FRIPS) Mass Spectrometry Results. 390, 110–117 (2015) (http://dx.doi.org/10.1016/j.ijms.2015.06.019) with permission from Elsevier, copyright 2015]

Full size image

In a previous publication, the different peptide fragmentation behavior in negative ion mode was understood to arise from the reversal of the relative energy order of the two CA processes [30]. In the positive ion mode, the first CA process (i.e., the radical generation step) required less internal energy than the second CA process (the radical-driven peptide backbone dissociations step) did. Therefore, after the application of the first CA, an additional internal energy needs to be supplied to induce peptide fragmentations through the application of the second CA process. In contrast, in negative ion mode, the first CA step required more internal energy than the second step did. Therefore, once sufficiently high energy was deposited in the negative ion mode to the radical precursor-containing peptide ions, the resulting radical peptide ions (even after the homolytic O–C bond cleavage) still possessed sufficient internal energy to carry out the next process of extensive peptide backbone dissociations.

As briefly indicated above, thus far the TEMPO-assisted FRIPS approach has been mostly used in ion-trap mass spectrometers, primarily because of the need for the two step CA applications. In the triple quadrupole mass spectrometer or hybrid quadrupole time-of-flight (Q-TOF) mass spectrometer, the two step CA issue may be circumvented by using in-source fragmentations (or nozzle-skimmer fragmentations) that can induce radical generation prior to entering the collision cell. However, because in-source fragmentations occur without the mass-selection of the precursor peptide ions of interest, this method cannot be the best method, particularly with the potential of using the TEMPO-assisted FRIPS approach in proteomics research.

In the current practical proteomics research, the hybrid Q-TOF and Orbitrap-based mass spectrometers are the two main work-horses, owing to their high mass resolving power [37–40]. In both instruments, collisional energy is relatively higher (i.e., 1–200 eV) than in the ion-trap instrument (1–20 eV) [38]. As suggested in the negative ion mode TEMPO-assisted FRIPS approach, if a collisional energy high enough to carry peptide backbone dissociations can be provided in the initial CA process in the positive ion mode, there is a good chance that peptide backbone dissociations can occur after a single application of collisional activation [30]. Because the significantly higher collision energy can be supplied in the collisional activation process in both Q-TOF and Orbitrap-based mass spectrometers, it is extremely worthwhile to evaluate how the TEMPO-assisted FRIPS takes place in these two instruments. Motivated by this reasoning, in the present study we carried out the TEMPO-assisted FRIPS experiments for a number of different model peptides in both Q-TOF and Orbitrap mass spectrometers and also compare these with the linear ion-trap mass spectrometer. Detailed results are described in the following sections.

Experimental

Materials

Bradykinin and cytochrome c from bovine heart were purchased from Sigma (St. Louis, MO, USA) and peptides H7405 (GRGLSLSR), H1990 (ALPMHIR), H2686 (YIYGSFK), H6190 (GRGDSPK), and N1120 (PHPFHFFVYK) were obtained from Bachem (Bubendorf, Switzerland). These peptides were used without further purification. o-TEMPO-Bz-C(O)-NHS was provided by Diatech Korea (Seoul, Korea). A sequencing grade modified trypsin was obtained from Promega (Madison, WI, USA). Water and acetonitrile of HPLC grade were procured from Burdick and Jackson (Ulsan, Korea). o-Methylisourea hemisulfate (OMIU), DL-dithiothreitol (DTT), iodoacetic acid (IAA), formic acid (FA), ammonium hydroxide, trifluoroacetic acid (TFA), dimethyl sulfoxide (DMSO), and triethylammonium bicarbonate (TEAB) were obtained from Sigma (St. Louis, MO, USA).

Sample Preparation

Cytochrome c (100 pmol) was digested as follows. Cytochrome c was reduced by incubation in 50 μM DTT for 10 min and was alkylated by reaction with 100 μM IAA for 10 min in the dark. The resulting cytochrome c solution was mixed with trypsin at a 1:50 (w/w) ratio, adjusted to pH 8–9 with 50 mM TEAB buffer and incubated overnight at 37 °C.

The peptides H2686 (YIYGSFK), H6190 (GRGDSPK), and N1120 (PHPFHFFVYK), and cytochrome c tryptic digests were guanidinated at the lysine side chain using a previously described procedure [36]. Blocking the lysine side chain primary amine using this guanidination procedure was necessary to avoid the conjugations of o-TEMPO-Bz-C(O)-NHS at both the lysine side chain and the wanted N-terminus. It was previously observed that the doubly conjugated peptides did not undergo the expected radical-driven peptide backbone fragmentations through FRIPS [36]. In the case of bradykinin and peptides H7405 (GRGLSLSR) and H1990 (ALPMHIR), this guanidination procedure was not necessary.

Both the guanidinated and unguanidinated peptides were subjected to o-TEMPO-Bz-C(O)-NHS conjugation. The detailed procedure is as follows. The dried peptides were dissolved in anhydrous DMSO solvent at a concentration of 100 μM. Separately, 1 mg of o-TEMPO-Bz-C(O)-NHS was dissolved in 100 μL anhydrous DMSO. Volumes of 5 μL of peptides in DMSO and 2 μL o-TEMPO-Bz-C(O)-NHS in DMSO were co-added to a total of 100 μL DMSO solution. After adjusting the pH to 8–9 with a 100 mM TEAB buffer, this solution was vortexed for 1 min and reacted for 6 h. The final o-TEMPO-Bz-C(O)-peptides were purified using a C18 reversed-phase solid phase extraction spin column (Ultra-micro spin column; Harvard Bioscience, Holliston, MA, USA) and dried using a SpeedVac.

Mass Spectrometry and LC-MS/MS

Tandem mass spectrometry experiments were carried out in positive ion mode on three mass spectrometers; a linear ion-trap (LTQ XL; Thermo Scientific, Bremen, Germany), a hybrid Q-TOF (Compact; Bruker Daltonics, Bremen, Germany), and a Q-Exactive hybrid quadrupole Orbitrap mass spectrometer (Thermo Scientific, Bremen, Germany). For the TEMPO-assisted FRIPS experiments, the peptides of interest were electrosprayed using syringe-pumped direct infusion. The peptide ions were isolated/filtered and subjected to MS3 on the linear ion-trap and MS/MS on the Q-TOF and Q-Exactive Orbitrap instruments.

In the linear ion-trap, the spray voltage was set to +1.8 to 2.0 kV, the other instrumental parameters were set as follows: capillary temperature, 275 °C; capillary voltage, +35 V; tube lens, +50 V; and isolation width, ±3 Da. The final MS3 mass spectra were acquired by averaging over 100 scans. In the hybrid Q-TOF mass spectrometer, the sample of interest was directly infused through an electrospray ionization source at a flow rate of 1 μL/min. The following instrument parameters were used: dry gas, 4.0 L/min; dry gas temperature, 200 °C; nebulizer pressure, 0.4 bar; capillary voltage, +4,000 V; and isolation width, ±3 Da. The final MS/MS spectra were obtained by averaging the MS/MS spectra over 1 min. In the Q-Exactive Orbitrap instrument, a higher energy collisional dissociation (HCD) tandem mass spectrometry method was used for peptide fragmentations. The following setup parameters were used: spray voltage, +2.0–2.5 kV; capillary temperature, 270 °C; S-lens, +55 V; and isolation width, ±3 Da. In the nano-ESI-LC-MS/MS experiments, to obtain HCD MS/MS spectra, a data-independent MS/MS scan (HCD) was performed at various normalized collision energy (NCE) values ranging from 10% to 30% with an isolation width of 5 Da. The mass resolution (Rmass) and maximum ion injection time (tmax) for full MS scans were set to 70,000 and 100 ms, respectively, and 35,000 and 60 ms were used for the MS/MS scan. A binary gradient elution was made using the model 1260 capillary LC system from Agilent Technologies. The samples were separated using a fused silica capillary column (150 mm × 75 μm i.d. × 360 μm o.d.) packed in-house with Magic C18 AQ (3 μm, 100 A) resins. The trap column, as an online sample concentrator, was packed in a 4 cm capillary (100 μm i.d., 360 μm o.d.) with Magic C18AQ (5 μm, 200 A) resins for 0.5 cm. For the RPLC runs, buffer A (2% ACN in water) and B (96% ACN with 2% DMSO in water), both with 0.1% formic acid, were used as mobile phases. The RPLC gradient conditions operated at 200 nL/min were as follows: the o-TEMPO-Bz-C(O)-incorporated peptides were injected into the trap column first by 2% B for 10 min via an autosampler, and initially increased to 10% B for 0.5 min, then linearly increased to 23% up to 24.5 min., to 40% up to 15 min, ramped to 85% for 3 min, held at 85% for 8 min to clean the RP column, and finally decreased to 2% over 2 min, and held for 18 min to re-equilibrate the RP column.

Results and Discussion

Q-TOF Results

A number of representative o-TEMPO-Bz-C(O) conjugated peptides were subjected to collisional activation in the platform of the hybrid Q-TOF mass spectrometer. As an example, Figure 1 shows the resulting collisional activation mass spectra of singly protonated o-TEMPO-Bz-C(O)-bradykinin (RPPGFSPFR), (RM + H)+ at m/z 1333.7, obtained at the collision energies of Figure 1a 40, Figure 1b 60, and Figure 1c 75 (arbitrary units in the Bruker Compact Q-TOF mass spectrometer), where the subscript R in (RM + H)+ indicates that the o-TEMPO-Bz-C(O)- group was conjugated to the N-terminus of bradykinin. At the collision energy of 40 (arbitrary unit), the radical intermediate •Bz-C(O)-bradykinin ions, (rM + H)+ at m/z 1177.6, were generated as a major product as a result of the homolytic cleavage of the bond between the benzylic carbon and the oxygen of the TEMPO moiety (Figure 1a) [8, 26, 30–36]. In addition, several peptide backbone fragments were also observed in the region of m/z 400–1000, though their abundances are very low. The observed peptide fragments included [z 4 ]+, [ r a 5 +H]+•, [ r c 5 +3H]+ •, [y 6 -H]+•, [ r a 6 +H]+•, [y 7 ]+, [x 7 +H]+•, [y 8 +], [x 8 +H]+ •, and [ r a 8 +H]+• in m/z increasing order, which contained a number of typical a/x- and c/z-type radical-driven peptide backbone fragments. (The nomenclature of radical fragment ions follows the rule described in reference [41]). Here, the subscript r, e.g., in [ r a 5 +H]+• indicates that the •Bz-C(O)- group is attached to the N-terminus of the corresponding fragments. In our previous TEMPO-assisted FRIPS MS studies, it was suggested that the nascent radical site of the •Bz-C(O)-bradykinin ions abstracted a hydrogen atom from the beta carbon (Cβ) of a side chain, leading to the migration of the radical site to the Cβ [8, 34]. The migrated radical site was thought to eventually lead to the production of a/x or c/z-type peptide backbone fragments through β-elimination. The dominant production of a/x or c/z-type backbone fragments were also observed in other radical-driven peptide backbone dissociation studies [3–5, 7, 10, 12].

Figure 1
figure 1

MS/MS mass spectra of singly protonated o-TEMPO-Bz-C(O)-bradykinin (RPPGFSPFR), (RM + H)+ at m/z 1333.7, using the hybrid Q-TOF mass spectrometer at the collision energy of (a) 40, (b) 60, and (c) 75 (arbitrary units)

Full size image

As the collision energy was adjusted to the higher collision energy of 60 (arbitrary units), a larger number of peptide backbone fragments as well as side-chain loss peaks appeared in a wide m/z region below m/z 1177.6, (rM + H)+, and their abundances significantly increased (Figure 1b). It was also observed that the relative abundances of the peptide fragments were slightly different from those in Figure 1a. For example, the relative abundances of [y 7 ]+ and [y 8 +] substantially increased, whereas the relative abundance of [ r c 5 + 3H]+• significantly reduced.

Production of a large number of peptide fragments in the hybrid Q-TOF collisional activation experiment, wherein only the single-step collisional activation was applied, is an unexpected result because in the previous ion-trap studies, two-stage ion-trap tandem mass spectrometry applications were required for o-TEMPO-Bz-C(O)-peptides to produce a large number of peptide backbone fragments in positive ion mode [30].

To see what happens at much higher collision energy, the collision energy of 75 (arbitrary units) was applied, and the resulting tandem mass spectrum is shown in Figure 1c. The obtained mass spectrum was quite similar to Figure 1b in that a large number of peptide backbone fragments were observed over a wide m/z region. However, it is notable in Figure 1c that the abundances of [y 7 ]+ and [y 8 +] were the highest among many peptide fragments. The abundant y-type fragments presumably arose from charge-driven tandem mass spectrometry processes rather than from radical-driven peptide fragmentation processes [31, 34, 42, 43]. The higher collisional energy appears to more widely open the charge-driven fragmentation channels in the dissociation dynamics. In general, the charge-driven peptide backbone fragmentation mechanism, which is often promoted by the so-called “mobile proton’, is in competition with the radical-driven dissociation pathways [31, 34, 42, 43]. When these two fragmentation pathways are possible, a major pathway can be determined depending on the total number of protons, cationization agent, and the presence of the most basic arginine residue in the sequence [14, 31, 34, 43, 44]. In the present study, the application of high collision energy [i.e., collision energy of 75 (arbitrary units)], made the thermodynamically unfavorable charge-driven pathway (leading to the production of [y 7 ]+ and [y 8 +] fragments) more accessible than in the low collision energy experiments (i.e., collision energies of 40 and 65).

The occurrence of extensive peptide backbone fragmentations, particularly those of radical character, at the single-stage tandem mass spectrometry application in the hybrid Q-TOF can be explained as follows. In the collision cell Q2 of the hybrid Q-TOF mass spectrometer, the two-step activation processes may take place sequentially, attributable to the multiple collisions in Q2 [37, 38]. Due to the multiple collisions at Q2, the initially generated •Bz-C(O)-bradykinin ions, (rM + H)+, which still possess high kinetic energy even after the homolytic cleavage, can experience additional collisional activation leading to peptide backbone dissociations. In contrast, the additional collisional activation cannot take place in the ion trap instrument. In the ion-trap instrument, collisional activation occurs through the resonant excitation of the ions at certain m/z values. Once the precursor ions undergo homolytic cleavage to generate •Bz-C(O)-bradykinin ions through resonance excitation, the m/z value of the generated •Bz-C(O)-bradykinin ions is not the original one any more, and therefore the •Bz-C(O)-bradykinin ions cannot be further resonantly excited within the same collisional activation step [38, 45]. Due to this reason, although the •Bz-C(O)-bradykinin ions generated in the ion trap instrument would experience multiple collisions as in the hybrid Q-TOF MS/MS experiments, they could not be collisionally activated sufficiently to induce peptide backbone dissociations.

Furthermore, higher collision energies at the Q2 stage (i.e., in the collisional energy range of 40–75 arbitrary units) may be used for carrying over the two sequential fragmentation processes in the collision cell. In the previous study performed in negative ion mode, the threshold energy for the second step (radical-driven peptide fragmentation) was lower than that of the first step (radical-generation) [30]. Thus, even a single-stage collisional activation by resonance-excitation in the ion-trap mass spectrometer was sufficient to induce radical-driven peptide backbone fragmentations to a substantial extent. Because a much higher collisional energy can be provided in the hybrid Q-TOF mass spectrometer, the two sequential processes (radical generation and radical-driven peptide fragmentations) can be carried out, even in positive ion mode, by the single application of MS/MS within the collision cell and can be detected.

These two processes mentioned above are likely to be at work together to result in extensive peptide backbone dissociations in the positive-ion hybrid Q-TOF tandem mass spectrometry experiments.

FRIPS by HCD (Higher-Energy Collisional Dissociations)

The TEMPO-assisted FRIPS experiments were also carried out in the Q-Exactive Orbitrap mass spectrometer; the use of which is now rapidly expanding [14, 39, 40, 44]. Indeed, the Sinz and Schäfer groups recently used the Orbitrap mass spectrometer in their studies with a bifunctional TEMPO-active ester reagent for peptide structure analysis by FRIPS MS [14, 44]. In the present experiment, HCD collisional activation was used.

Figure 2 shows the resulting tandem mass spectra for singly protonated o-TEMPO-Bz-C(O)-bradykinin (RPPGFSPFR), (RM + H)+ at m/z 1333.7, which was obtained using the HCD method at the NCE of Figure 2a, 14, Figure 2b, 20, and Figure 2c, 27 (arbitrary units). Most importantly, as for the Q-TOF experiments, a large number of peptide backbone fragment ions could be detected in the HCD applications, although only a single collisional application was made. In particular, a large number of typical a/x- and c/z-type radical-driven peptide backbone fragments as well as some minor b/y-type charge-driven peptide backbone fragments were observed: [z 4 ]+, [ r a 5 +H]+•, [ r c 5 +3H]+ •, [y 6 – H]+•, [ r a 6 +H]+•, [y 7 ]+, [x 7 +H]+•, [y 8 +], [x 8 +H]+ •, and [ r a 8 +H]+• [41]. The observation of a large number of a/x- and c/z-type fragments indicates that in the HCD MS/MS of the Orbitrap mass spectrometer, the radical-driven peptide backbone dissociation mechanism is also the dominant peptide fragmentation pathway. In addition, it is notable that the mass spectra shown in Figure 2 are very similar to those in Figure 1 in terms of the relative abundances of the observed fragments. Specifically, as the collision energy increased, the relative abundances for y-type ions (e.g., [y 7 ]+ at m/z 805.40 and [y 8 +] at m/z 902.45) gradually increased, whereas the abundances for c-type ions (e.g., [ r c 5 +3H]+• at m/z 690.37) decreased. As explained above, the higher collision energy made the charge-driven peptide backbone dissociation pathway, which led to the production of [y 7 ]+ and [y 8 +] fragments, more accessible.

Figure 2
figure 2

HCD MS/MS mass spectra of singly protonated o-TEMPO-Bz-C(O)-bradykinin (RPPGFSPFR), (RM + H)+ at m/z 1333.7, using Q-Exactive Orbitrap instrument at the normalized collision energy of (a) 14, (b) 20, and (c) 27 (arbitrary units)

Full size image

Considering that the collisional activation in both the hybrid Q-TOF instrument and Q-Exactive are produced in a similar manner, the similarity of the Q-TOF and Q-Exactive Orbitrap results can be easily understood. In these two methods, the precursor ions are accelerated within the collision cell through the potential difference between the two adjacent compartment electrodes [37–40].

Comparison of FRIPS MS Results of Ion-Trap, Q-TOF, and Orbitrap Mass Spectrometers

Here, to compare the peptide fragmentation behaviors occurring under the conditions of low energy collisions and higher-energy collisions, the results from the linear ion-trap, Q-TOF, and Q-Exactive Orbitrap mass spectrometers are compared. In Figure 3a, ion-trap MS3 spectrum, Figure 3b, Q-TOF MS/MS spectrum (at the collision energy of 75), and Figure 3c, Orbitrap HCD MS/MS spectrum (at the collision energy of 27 NCE) in positive ion mode for singly protonated o-TEMPO-Bz-C(O)-bradykinin, (RM + H)+ m/z 1333.7, are shown.

Figure 3
figure 3

Comparison of (a) MS3 ion-trap mass spectrum, (b) MS/MS mass spectrum on the hybrid Q-TOF instrument (75, arbitrary units), and (c) HCD MS/MS mass spectrum on the Q-Exactive instrument (27, arbitrary units), acquired for singly protonated o-TEMPO-Bz-C(O)-bradykinin (RPPGFSPFR), (RM + H)+ at m/z 1333.7

Full size image

The most striking difference between the low and higher collision energy results is that the ion-trap experiments required two distinct applications of collisional activation for the observation of a large number of peptide backbone fragments, whereas a single application of collisional activation was sufficient for substantial peptide backbone fragmentations in the Q-TOF and Q-Exactive Orbitrap experiments. In the ion-trap experiments, only low collision energies (i.e., 1–20 eV) can be provided to the precursor ions of interest, compared with higher collisional energies in the range of 1–200 eV in the hybrid Q-TOF and Q-Exactive Orbitrap instruments [38]. The higher collisional energies, which can be provided in the Q-TOF and Q-Exactive Orbitrap instruments, enable the two ensuing TEMPO-assisted FRIPS peptide backbone dissociations steps (i.e., the radical generation step and the radical-driven peptide backbone dissociation step) to occur in a single collisional activation event. Furthermore, the resonant excitation mechanism of the ion-trap instrument prevented the once-generated •Bz-C(O)-peptide ions from being further activated in the ion trap [38, 45]. Thus, an additional collisional activation step needed to be supplied to the •Bz-C(O)-peptide ions.

On the other hand, in terms of the relative abundances of the peptide fragment types, in the ion-trap spectrum, the abundances of a- and c-type ions were high and the abundances of y-type ions were relatively low. However, for both Q-TOF and Orbitrap MS/MS experiments, the relative abundances of y-type fragments, particularly [y 7 +] and [y 8 +], significantly increased. It is also notable that a few additional side-chain loss peaks accompanying peptide backbone fragments (e.g., [y 7 – 44 + 3H]+• and [ r b 7 + – 86]+•) additionally appeared in the Q-TOF and the HCD MS/MS spectra. The increased relative abundances of the y-type fragments in the Q-TOF and Orbitrap experiments can be ascribed to the increased relative contribution of the charge-driven peptide fragmentation pathways with respect to the radical-driven dissociation pathways through higher collision energies in the Q-TOF and Orbitrap instruments [31, 34, 42, 43].

The above-mentioned observations were similarly observed for the other peptides, such as H7405, H1990, H2686, H6190, and N1120, the spectra of which are shown in Supplementary Figures 1–5, respectively. That is, extensive peptide fragmentations were observed at a single application of MS/MS in both Q-TOF and Q-Exactive Orbitrap instruments, and the fragmentation patterns were more or less similar to each other and also to the MS3 spectrum from the ion-trap instrument.

LC-MS/MS Experiments Using the Q-Exactive Instrument

To see whether the one-step collisional activation experiments could induce extensive backbone fragmentations for the o-TEMPO-Bz-CO conjugated peptides eluted from the reversed-phase chromatography column, nano-ESI-LC-MS/MS experiments using the Q-Exactive Orbitrap instrument were conducted using the peptide mixture prepared by the sequential guanidination and o-TEMPO-Bz-CO- conjugation for the tryptic digests of cytochrome c. Figure 4 shows the acquired base-peak chromatogram (BPC). In general, the conjugated peptides appeared in the later chromatographic retention times compared with the unconjugated peptides. The hydrophobic nature of the o-TEMPO-Bz-CO- conjugated group substantially delayed the retention time of the conjugated peptides in the reversed-phase chromatogram. Analysis of the chromatogram revealed that the conjugation efficiency could be further improved (Supplementary Table 1).

Figure 4
figure 4

Nano-ESI-LC-MS base-peak chromatogram (BPC) obtained for the peptide mixture prepared using the sequential sample preparation of guanidination and o-TEMPO-Bz-CO conjugation of the tryptic digests of cytochrome c

Full size image

Figure 5 shows some examples of LC-MS/MS mass spectra acquired for Figure 5a, EDLIAYLKG (C7 in the BPC), Figure 5b, MIFAGIKG (C28 in the BPC), and Figure 5c, TGPNLHGLFGR (C34 in the BPC), wherein the subscript G indicates that lysine side chain was guanidinated. As clearly illustrated in Figure 5, a single-step HCD produced extensive peptide backbone fragmentations, including radical-driven fragments in a major portion and some charge-driven fragments. These results demonstrate that the TEMPO-assisted FRIPS approach is applicable in the nano-LC-MS/MS platforms that utilize the hybrid Q-TOF and Q-Exactive Orbitrap instruments, although improvements in the conjugation efficiency are needed.

Figure 5
figure 5

HCD LC-MS/MS mass spectra of (a) EDLIAYLK (C7 in the BPC), (b) MIFAGIK (C28 in the BPC), and (c) TGPNLHGLFGR (C34 in the BPC) from a tryptic digestion mixture of cytochrome c

Full size image

Conclusions

In the previous ion-trap studies, it was shown that two-step applications of collisional activation were necessary to observe radical-driven peptide fragmentations in the positive-ion TEMPO-assisted FRIPS experiments. The issue of the two-step collisional activation processes needs to be addressed in the TEMPO-assisted FRIPS MS because the sensitivity of MS3 is generally lower than that of MS/MS and, more importantly, the duty-cycle is lower. At a first glance, it appeared that the TEMPO-assisted FRIPS MS was difficult to implement in the Q-TOF and Q-Exactive instruments because these mass spectrometers are known to be equipped only for a single-stage tandem mass spectrometry method. Therefore, it was hypothesized that by applying “in-source fragmentations,” two-step applications of collisional activation could be implementable. Unexpectedly, however, one-step peptide backbone fragmentations are possible in TEMPO-assisted FRIPS using Q-TOF MS/MS and Q-Exactive Orbitrap HCD MS/MS because of their unique higher energy deposition and the additional multiple collisions in the collision cell. This finding clearly demonstrates that the TEMPO-assisted FRIPS approach appears to become more attractive in peptide mass spectrometry, particularly in the platform of the hybrid Q-TOF and Q-Exactive Orbitrap instruments. With a few more improvements with regard to solubility, conjugation efficiency, and one-step peptide fragmentations in even the linear ion-trap instruments, the TEMPO-assisted FRIPS approach is expected to be a very useful tool in peptide mass spectrometry research.